lable at ScienceDirect

Environmental Pollution 157 (2009) 331–340

Contents lists avai

Environmental Pollution

journal homepage: www.elsevier .com/locate/envpol

Hg concentrations and related risk assessment in coral reefcrustaceans, molluscs and fish from New Caledonia

Tiphaine Chouvelon a, Michel Warnau b,1, Carine Churlaud c, Paco Bustamante a,*

a Littoral Environnement et Societes (LIENSs), UMR 6250, CNRS-Universite La Rochelle, 2 Rue Olympe de Gouges, F-17042 La Rochelle Cedex 01, Franceb International Atomic Energy Agency-Marine Environment Laboratories (IAEA-MEL), 4 Quai Antoine Ier, MC-98000, Monacoc Federation de Recherche en Environnement pour le Developpement Durable (FR-EDD), FR CNRS 3097, Centre Commun d’Analyses,5 Perspectives de l’Ocean, F-17071 La Rochelle Cedex 9, France

This work reports the first assessment of Hg levels in edible organism

s from the New Caledonian lagoonand the associated risk linked to their consumption by Human.

a r t i c l e i n f o

Article history:Received 8 April 2008Received in revised form 15 June 2008Accepted 17 June 2008

Keywords:Tropical environmentMetalMarine productsSeafoodRisk assessmentPTWI

* Corresponding author. Tel.: þ33 5 46 50 76 25; faE-mail address: pbustama@univ-lr.fr (P. Bustaman

1 Present address: LIENSs, UMR 6250 CNRS-UniveOlympe de Gouges, F-17042 La Rochelle, France.

0269-7491/$ – see front matter � 2008 Elsevier Ltd.doi:10.1016/j.envpol.2008.06.027

a b s t r a c t

There is a dramatic lack of data on Hg levels in marine organisms from tropical areas, and in particularfrom New Caledonia. For the first time, this study reports the total Hg concentrations in the tissues ofseveral marine taxa from the New Caledonian lagoon. Seafood from both wild and farmed populationswas considered. Hg concentrations varied over three orders of magnitudes according to factors includingspecies, age (size/weight), trophic level, lifestyle and geographical origin. Taking into account the edibletissues, estimations of the amount of flesh that should be consumed by a 60-kg person to reach the HgProvisional Tolerable Weekly Intake (PTWI) reveal acceptable risk for Human health in general. However,a risk was clearly identified in one site of the lagoon (i.e. Grande Rade) where high Hg concentrationswere measured. These concentrations were higher than values reported in the current literature.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction

Among heavy metals, Hg is of particular concern in marineenvironmental studies because it has both natural and anthropo-genic sources, has no known biological function and is toxic to allliving organisms (for a review, see Eisler, 1987). In the marine en-vironment, Hg is naturally released in the water and then in theatmosphere through erosion of the sediments where it is present assulphide salts (Salomons et al., 1987). However, about 30% of thetotal Hg emissions in the atmosphere are of anthropogenic origins.Moreover, these anthropogenic atmospheric emissions are 10 timesgreater than direct inputs to water, contributing to a widespreadcontamination, which is difficult to assess (Hylander, 2001). Theseemissions are mainly originating from industrial and domestic in-cineration (Cossa et al., 1990).

In the marine environment, Hg is mainly found under organicforms. Among organic Hg species, methyl-Hg is the most stableform in the marine environment; it is primarily produced by

x: þ33 5 46 45 62 84.te).rsite de La Rochelle, 2 Rue

All rights reserved.

microorganisms in sediments. Methyl-Hg is also the most toxicform to organisms (WHO, 1990). Finally, Hg and methyl-Hg havethe particularity to be bioaccumulated by marine organisms and tobiomagnify along the food chain (Zizek et al., 2007). Therefore, theconsumption of marine products represents a non-negligible ex-posure pathway to Hg and, thereby, a risk for Human health (e.g.Buzina et al., 1989; Svensson et al., 1992). This is particularly thecase for fish, since virtually 100% of the total Hg in fish muscles(edible tissues) are present as methyl-Hg (e.g. Bloom, 1992; Hols-beek et al., 1997), and is thus highly bioavailable for Human con-sumers and other predators.

In New Caledonia, coastal waters are subjected to large inputs ofmetals, mainly due to intense mining activities (especially Ni pro-duction), but also due to natural erosion of the soils associated withtropical rainfall, urban development and lack of efficient waste-water treatment (Ambastian et al., 1997). Surprisingly, few studieshave been conducted to assess the marine contamination status ofNew Caledonia by metals (Labrosse et al., 2000). Available studiesare generally limited to Ni and its mining by-products and toa narrow range of species (Monniot et al., 1994; Bustamante et al.,2000; Hedouin et al., 2007, in press; Metian et al., 2008). To the bestof our knowledge, data published on Hg in marine organisms fromNew Caledonia are limited to a single study on pelagic marine

T. Chouvelon et al. / Environmental Pollution 157 (2009) 331–340332

mammals (Bustamante et al., 2003), and so far none is available onnon-mammalian organisms from the lagoon.

The aim of this study was therefore to provide baseline in-formation on the Hg contamination status of the New Caledoniacoastal marine environment. A wide range of species includingBivalves, Cephalopods and Teleosts from different locations wereanalysed for their Hg concentrations. Special emphasis was given tospecies that have already been recognised as valuable bioindicatorspecies (i.e. the tumid venus Gafrarium tumidum and the radulascallop Comptopallium radula; Hedouin et al., 2007, in press; Metianet al., 2008), and to those that are major seafood for local con-sumption or exportation. The investigated organisms were col-lected either from the wild (i.e. scallops, clams, cockles,cephalopods and fish) or from mariculture farms (e.g. the Pacificblue shrimp Litopenaeus stylirostris and the Pacific cupped oysterCrassostrea gigas). Differences in Hg bioaccumulation among bodycompartments of a given species, within and among taxonomicgroups, and among sites were tested. Size and weight were alsotaken into account as driving parameters. Finally, this paper pro-vides a site-specific preliminary risk assessment for Hg consump-tion along with seafood.

2. Materials and methods

2.1. Sampling and sample preparation

Organisms belonging to 32 taxa (Table 1) were collected in March and October2007 along the coast of New Caledonia (Fig. 1, Table 2), either by SCUBA diving (fish),by hand picking at low tide (bivalves), or bought on the central market of NoumeaCity (shrimp, oysters, cephalopods and some fish). The sampling sites were selectedbecause they were reported to have contrasting contamination status (Hedouinet al., in press; Metian et al., 2008; Warnau and Bustamante, unpublished results):Grande Rade, Grande Rade-SLN, Koutio Bay, the edge of Noumea harbour, SainteMarie Bay, Ouano Bay (which included three sub-sampling sites from the shore – S1 –to open waters – S3), Maa Bay, Unia and Tomo (Fig. 1, Table 2). The geographicalorigin of the organisms bought on the Noumea central market was obtained fromthe selling fishermen: Bassin de La Foa, Dumbea Bay, southern and northern lagoon(Fig. 1, Table 2).

Grande Rade is subjected to anthropogenic inputs from the Ducos industrialzone, whereas Grande Rade-SLN is directly subjected to those of the metallurgicfactory ‘‘Societe Le Nickel’’ (SLN). Due to these anthropogenic inputs, Grande Radesediments were shown to display high concentrations in several elements (Hedouinet al., in press). Koutio Bay is characterised by the presence of an important rubbishdump and is influenced by inputs of domestic wastes from Noumea City. SainteMarie Bay is located to the East of Noumea City and receives important sewagesludge and terrigenous inputs coming from the Coulee River. Dumbea Bay is alsosubjected to terrigenous inputs from the Dumbea River (Ambastian et al., 1997). Incontrast, Unia (East coast), southern and northern lagoon, Maa Bay and Ouano Bayare preserved from important anthropogenic inputs. Finally, the Tomo site is locatednear the international airport of New Caledonia, between Ouano Bay and NoumeaCity, whereas the site of Bassin de La Foa is located near a shrimp farm on the Westcoast of New Caledonia.

All collected organisms were weighed (wet wt) and measured (total length forfish and shrimp, height and length of the shells for bivalves, and mantle length forcephalopods) upon return to the laboratory. Characteristics (origin, number of in-dividuals, length and weight, sampling date) of each of the 32 species collected aregiven in Table 1. This table also indicates the trophic level (i.e. filter feeder, grazer/scavenger, predator of invertebrates, predator of invertebrates and small fish,predator of small fish) and the water-column distribution (benthic, nectobenthicand neritic) for every species.

Organisms were dissected to collect specific body compartments. A piece ofmuscle (edible tissue) was taken from fish, shrimp and cephalopods. For fish, theliver was also collected. In cephalopods (Sepioteuthis lessoniana, Sepia latimanus andOctopus cyanea), the digestive gland and branchial hearts were collected wheneverpossible. In the case of Pectinidae (C. radula and Mimachlamys gloriosa), the edibletissues (muscleþ gonads) were separated from the remaining soft tissues. For allother bivalves, the soft parts were considered as a whole. After dissection, thesamples were weighed (wet wt) and immediately placed in individual plastic bags,frozen at �25 �C, freeze-dried, and weighed again (dry wt). Freeze-dried tissueswere then ground and stored in individual plastic vials until further metal analysis.

2.2. Total Hg analysis

Total Hg analyses were carried out with an Advanced Mercury Analyser (ALTECAMA 254), on dried tissue aliquots ranging from 4 to 50 mg, weighed to the nearest

0.01 mg. For Hg determination, the metal was evaporated by progressive heating upto 800 �C, then held under oxygen atmosphere for 3 min, and finally amalgamatedon a gold net. Afterwards, the net was heated to liberate the collected Hg, which wasmeasured by atomic absorption spectrophotometry. Mercury analyses were runaccording to a thorough quality control programme including the analysis of a NRCreference material (lobster hepatopancreas TORT-2; National Research Council,Canada). Standard aliquots were treated and analysed according to the same con-ditions as the samples. The results were in good agreement with the certified values,with an average recovery rate of 105%. Detection limit was 5 ng g�1 dry wt. All totalHg concentrations in tissues further reported are expressed in ng g�1 dry wt.

2.3. Data analysis

All data submitted to statistical tests were first checked for normality (Shapiro–Wilk test) and for homogeneity of variances (Bartlett test). When these conditionswere satisfied, parametric tests were used in the subsequent analyses; otherwise,non-parametric analogues were used. Pearson or Spearman correlation coefficienttest was used to analyse the correlations between size or weight and total Hgconcentration in edible tissues in each species. Differences in Hg concentrationsbetween tissues in species for which different body compartments were dissectedwere tested by Student’s t-test or Wilcoxon test for paired samples. To assesswhether trophic levels (see Table 1) could influence Hg concentrations in fish(among the collected organisms, this taxa was the only one to display different diets;Table 1), analysis of covariance (ANCOVA) was performed, using length of in-dividuals as covariable. Finally, species with a minimum of two individuals collectedper different location were considered to test differences in Hg concentrationsamong sampling locations (Table 4), using a one-way analysis of variance (ANOVA)followed by the post hoc Tukey test. In the case of species for which a correlationbetween size or weight and Hg concentration was previously revealed, ANCOVA wasperformed instead of ANOVA, using size or weight as covariable. One-way ANOVAand ANCOVA were performed on log-transformed Hg concentrations and length orweight values. When appropriate, the variability explained by each factor and theirinteraction was derived from the sum of squares (Warnau et al., 1996, 1998). Thelevels of significance for statistical analyses were always set at a¼ 0.05.

2.4. Risk assessment for Human consumers

To prevent Human health impairments by Hg originating from dietary sources,the Joint FAO/WHO Expert Committee on Food Additives (JECFA) allocated a Pro-visional Tolerable Weekly Intake (PTWI) for total Hg and methyl-Hg of 5000 and1600 ng kg�1 wk�1, respectively (WHO, 2003). In this study, we calculated a ‘‘Maxi-mum Safe Weekly Consumption’’ (MSWC), i.e. the weight of edible tissue (muscle forshrimp, cephalopods and fish, muscleþ gonads or whole soft parts for bivalves) thatshould be consumed by a 60 kg adult to reach the PTWI. For this, mean Hg con-centrations in ng g�1 dry wt were first converted to wet wt, taking into account thedry wt/wet wt ratios measured in all samples. These ratios indicated a meanmoisture content of 80% for bivalve flesh, shrimp and cephalopod muscle, and of 75%for fish muscle, which matched well those previously published for the same tax-onomic groups (e.g. Trombini et al., 2003; Bustamante and Miramand, 2005). Then,the PTWI (ng kg�1 wk�1) multiplied by the consumer body weight (60 kg) was di-vided by the Hg concentration (ng g�1 wet wt) in the considered seafood to obtainthe MSWC (g wet wt). The latter value (g) was finally converted in kg (Tables 3 and4). In the particular cases of the fish Plectropomus leopardus and the bivalves C.radula, G. tumidum and Anadara scapha, the MSWC was also calculated for each givensite (Table 4). Finally, the MSWC was calculated for total Hg and methyl-Hg, takinga 75% ratio of Me-Hg/T-Hg for shrimp muscle (Riisgard and Famme, 1986), an 80%ratio of Me-Hg/T-Hg for cephalopod muscle (Bustamante et al., 2006), and a 100%ratio for fish muscle (Bloom, 1992). MSWC for methyl-Hg was not calculated in bi-valves, due to the great variability of the ratio Me-Hg/T-Hg found in the literature(ranging from 20 to 95%), depending on species, age, season and location (Moh-lenberg and Riisgard, 1988; Kawaguchi et al., 1999; Otchere et al., 2003; Trombiniet al., 2003).

3. Results

3.1. Concentrations and tissue distribution

Total Hg concentrations in edible tissues varied greatly withinand among species groups (i.e. fish, crustaceans, mollusc cephalo-pods, and mollusc bivalves). They ranged from 26 to2063 ng g�1 dry wt in fish, from 128 to 297 ng g�1 dry wt in crus-taceans (i.e. L. stylirostris), from 41 to 218 ng g�1 dry wt in cepha-lopods, and from 73 to 2531 ng g�1 dry wt in bivalves (see Table 3).Even within a single species, the variability was sometimes quiteimportant, e.g. for Periglypta chemnitzi, Hg concentrations rangingfrom 244 to 890 ng g�1 dry wt. As it could be expected the vari-ability was even greater when animals of a single species were

Table 1Characteristics of the organisms collected and analysed

Family and Scientific Name Common Name n Length (mm) Mean� SD(Range)

Wet wt (g) Mean� SD(Range)

Samplingdate

Sampling site(s) Trophic level Life style

CrustaceansPenaeidae

Litopenaeus stylirostris Pacific blue shrimp 5 155� 4 (155–160) 21.5� 1.3 (20.2–23.3) Mar 2007 Bassin de La Foa Grazer/Scavenger Benthic

Mollusc BivalvesArcidae

Anadara scapha Antique ark 28 58.4� 8.8 (28–70) 60.8� 20.5 (9–101) Mar 2007 Grande Rade, Koutio,Ouano (S1, S2), Maa,Sainte Marie

Filter feeder Benthic

Oct 2007 Grande Rade-SLNOstreidae

Crassostrea gigas Pacific cupped oyster 5 86.0� 4.7 (80–93) 62.0� 7.1 (54–73) Mar 2007 Dumbea Filter feeder BenthicPectinidae

Comptopallium radula Radula scallop 24 77.3� 14.3 (44–99) 70.6� 28.4 (19.6–115) Mar 2007 Grande Rade, Koutio,Maa, Sainte Marie

Filter feeder Benthic

Oct 2007 Noumea harbourMimachlamys gloriosa Tropical scallop 10 72.6� 6.4 (63–80) 55.8� 14.9 (38–80) Oct 2007 Noumea harbour Filter feeder Benthic

VeneridaeGafrarium tumidum Tumid venus 89 37.6� 4.6 (29–45) 22.3� 6.4 (11.3–35.8) Mar 2007 Grande Rade,

Grande Rade-SLN,Koutio, Ouano(S1, S2, S3), Maa

Filter feeder Benthic

Oct 2007 Unia, TomoPeriglypta chemnitzi Chemnitz Venus 10 60.6� 2.8 (56–65) 111� 26.3 (73–172) Oct 2007 Unia Filter feeder Benthic

Mollusc CephalopodsLoliginidae

Sepioteuthis lessoniana Big fin reef squid 9 215� 34a 441� 185 Oct 2007 Southern Lagoon Predator NeriticOctopodidae

Octopus cyanea Big blue octopus 7 113� 16.3 (82–131)a 1521� 671 (700–2900) Oct 2007 Northern Lagoon Predator BenthicSepiidae

Sepia latimanus Broad club cuttlefish 1 240a 1300a Oct 2007 Southern Lagoon Predator Nectobenthic

FishAcanthuridae

Naso unicornis Bluespine unicornfish 1 435 1600 Mar 2007 Ouano (S3) Grazer/Scavenger NeriticHaemulidae

Plectorhinchus albovittatus Two-striped sweetlips 1 405 1300 Mar 2007 Ouano (S3) Predator (invertebrates) NeriticP. chaetodonoides Harlequin sweetlips 1 351 900 Mar 2007 Grande Rade Predator (invertebrates and small fish) NeriticP. flavomaculatus Lemon sweetlips 1 485 1600 Mar 2007 Grande Rade Predator (crustaceans and small fish) NeriticDiagramma pictum Painted sweetlips 1 451 1160 Oct 2007 Ouano (S3) Predator (invertebrates and small fish) Neritic

KyphosidaeKyphosus vaigiensis Brassy chub 5 317� 50 (260–370) 718� 330 (380–1100) Mar 2007 Ouano (S3) Grazer/Scavenger Neritic

LabridaeBodianus perditio Golden-spot hogfish 1 343 800 Mar 2007 Grande Rade Predator (invertebrates) NeriticCheilinus chlorourus Floral wrasse 1 290 450 Mar 2007 Koutio Predator (invertebrates) Neritic

LethrinidaeLethrinus laticaudis Grass emperor 3 266� 11 (256–277) 376� 45 (331–420) Mar 2007 Southern Lagoon Predator (crustaceans and small fish) NeriticMonotaxis grandoculis Humpnose big-eye bream 2 261� 7 (256–266) 420� 21 (405–434) Mar 2007 Southern Lagoon Predator (invertebrates) Neritic

LutjanidaeLutjanus argentimaculatus Mangrove red snapper 5 449� 63 (348–510) 1345� 494 (625–1900) Mar 2007 Grande Rade,

Ouano (S3), MaaPredator (crustaceans andsmall fish)

Neritic

L. monostigma Onespot snapper 2 249� 37 (223–275) 475� 177 (350–600) Mar 2007 Ouano (S3), Maa Predator (crustaceansand small fish)

Neritic

(continued on next page)

T.Chouvelonet

al./Environm

entalPollution

157(2009)

331–340333

Tab

le1

(con

tin

ued

)

Fam

ily

and

Scie

nti

fic

Nam

eC

omm

onN

ame

nLe

ngt

h(m

m)

Mea

n�

SD(R

ange

)W

etw

t(g

)M

ean�

SD(R

ange

)Sa

mp

lin

gd

ate

Sam

pli

ng

site

(s)

Trop

hic

leve

lLi

fest

yle

Plat

ycep

hal

idae

Cym

bace

phal

us

beau

fort

iC

roco

dil

efi

sh2

510�

99

(44

0–

58

0)

142

5�

53

0(1

05

0–

180

0)

Oct

20

07So

uth

ern

Lago

onPr

edat

or(s

mal

lfi

sh)

Ben

thic

Pria

can

thid

aePr

iaca

nth

us

ham

rur

Moo

nta

ilb

ull

seye

73

07�

21(2

85

–3

40

)4

54�

96

(36

0–

60

0)

Mar

20

07O

uan

o(S

3)

Pred

ator

(in

vert

ebra

tes

and

smal

lfi

sh)

Ner

itic

Scar

idae

Scar

us

ghob

ban

Blu

e-b

arre

dp

arro

tfish

124

724

5M

ar2

007

Kou

tio

Gra

zer/

Scav

enge

rN

erit

icS.

mic

rorh

inos

Blu

nt-

hea

dp

arro

tfish

15

08

30

00

Mar

20

07O

uan

o(S

3)

Gra

zer/

Scav

enge

rN

erit

icS.

rivu

latu

sR

ivu

late

dp

arro

tfish

135

59

00

Mar

20

07O

uan

o(S

3)

Gra

zer/

Scav

enge

rN

erit

icS.

sch

lege

liYe

llow

ban

dp

arro

tfish

124

94

00

Mar

20

07O

uan

o(S

3)

Gra

zer/

Scav

enge

rN

erit

icSe

rran

idae

Plec

trop

omu

sle

opar

dus

Leop

ard

cora

lgro

up

er21

413�

110

(26

5–

615

)12

94�

1051

(30

0–

38

00

)M

ar2

007

Gra

nd

eR

ade,

Kou

tio,

Ou

ano

(S3

),M

aa,S

ain

teM

arie

Pred

ator

(sm

all

fish

)N

erit

ic

Oct

20

07G

ran

de

Rad

e,O

uan

o(S

3),

Sou

ther

nLa

goon

Epin

ephe

lus

coer

ule

opu

nct

atu

sW

hit

esp

otte

dgr

oup

er1

470

150

0O

ct2

007

Sou

ther

nLa

goon

Pred

ator

(cru

stac

ean

san

dsm

all

fish

)N

erit

ic

E.m

acu

latu

sH

igh

fin

grou

per

13

40

48

0M

ar2

007

Ou

ano

(S3

)Pr

edat

or(i

nve

rteb

rate

san

dsm

all

fish

)N

erit

icSp

arid

aeA

can

thop

agru

sbe

rda

Picn

icse

abre

am1

278

60

0M

ar2

007

Ou

ano

(S2

)Pr

edat

or(i

nve

rteb

rate

san

dsm

all

fish

)N

erit

ic1

271

50

0O

ct2

007

Maa

aM

antl

ele

ngt

h(n

otto

tal

len

gth

).

T. Chouvelon et al. / Environmental Pollution 157 (2009) 331–340334

sampled in different sites. Indeed, Hg concentrations ranged from173 to 2063 ng g�1 dry wt in P. leopardus, from 73 to 1015 ng g�1 drywt in A. scapha, and from 104 to 2531 ng g�1 dry wt in G. tumidum.In a single sampling location, Hg concentrations varied greatlyamong species of the same taxonomic group. For example inGrande Rade, average Hg concentrations were significantly differ-ent (Kruskal–Wallis test, p< 0.01) among the three bivalves C.radula, A. scapha and G. tumidum: 234� 24, 723� 414 and1633� 565 ng g�1 dry wt, respectively (Table 4). Interestingly, thispattern C. radula< A. scapha<G. tumidum was also observed inKoutio and Maa bays, even though the differences among specieswere significant only in Maa Bay (Kruskal–Wallis test, p< 0.01).

In Ouano Bay (S3), average Hg concentrations in the fish muscleswere significantly different (Kruskal–Wallis test, p< 0.01) amongKyphosus vaigiensis, Priacanthus hamrur and P. leopardus, with55� 22,188� 58 and 856� 330 ng g�1 dry wt, respectively (Tables 3and 4).

With very few exceptions, total Hg concentrations in fish werealways higher in the liver (detoxification organ) than in the muscle(edible tissue) (Fig. 2). Moreover, this difference was statisticallysignificant for most of the species (t or Wilcoxon tests for pairedsamples, p< 0.05). In contrast, in cephalopods, there was no sig-nificant difference in Hg concentrations between muscles and de-toxification tissues, i.e. the digestive gland for the squid S.lessoniana (t-test for paired samples, p> 0.05) and the branchialhearts for the octopus O. cyanea (Wilcoxon test for paired samples,p> 0.05) (Fig. 3). However, in this latter species, the concentrationsrecorded in the digestive gland of one specimen was far higherthose of the branchial hearts and muscle (Fig. 3). In S. latimanus, Hgconcentrations followed the decreasing order: branchial hearts>digestive gland>muscle (Fig. 3). Finally, for the Pectinidae (i.e. C.radula and M. gloriosa), Hg concentrations were significantly lower(Wilcoxon test for paired samples, p< 0.05) in the edible tissues(muscleþ gonads) than in the remaining tissues (Fig. 4).

3.2. Influence of size, weight and trophic level

Correlations between Hg concentrations in edible tissues andindividual size or weight were established for each species, con-sidering all sites together. Few Spearman correlations were signif-icant and Hg levels only varied positively with size for 2 out of the32 species collected: in the muscles of the fish P. leopardus(r¼ 0.773, p< 0.001) and in the muscleþ gonad of the pectinid C.radula (r¼ 0.422, p¼ 0.040). Interestingly, Hg concentrations inwhole soft parts of C. radula were also correlated significantly withsize (r¼ 0.472, p¼ 0.020). Positive correlations with the individualweight were significant for 3 out of the 32 species: in the muscles ofP. leopardus (r¼ 0.782, p< 0.001), and in the whole soft tissues ofthe bivalves G. tumidum and Periglypta chemnitzi (r¼ 0.239,p¼ 0.024 and r¼ 0.663, p¼ 0.037, respectively).

Differences in Hg concentrations among trophic levels in fish areshown in Fig. 5. The ANCOVA revealed that trophic level, length andtheir interaction were all factors affecting significantly Hg con-centrations (p< 0.005). Trophic level was the most important fac-tor explaining ca. 60% of Hg concentration variability, followed bylength (15%) and their interaction (6%). The remaining 19% of thevariability were due to other factors (residual term). It was notpossible to perform ANCOVA with the sites or species as covariable,as was done for trophic levels, because the residuals did not satisfynormality and/or homoscedasticity pre-requisites.

3.3. Differences among sampling sites

Table 4 displays the comparison among Hg concentrationsmeasured in different sites of New Caledonia for four species, i.e. thefish P. leopardus and the bivalves A. scapha, C. radula and G. tumidum.

MAA

OUANO

165°50'E 166°00'E 166°10'E 166°20'E 166°30'E 166°40'E

21°52'S

22°02'S

22°12'S

22°22'S

TOMO UNIA

SAINTE

MARIE

DUMBEA

LA FOA

KOUTIO

HARBOUR

GR

GR-SLN

NOUMEA

Fig. 1. Map of the sampling stations in the South of New Caledonia and in the vicinity of Noumea City. GR: Grande Rade; GR-SLN: Grande Rade-SLN.

T. Chouvelon et al. / Environmental Pollution 157 (2009) 331–340 335

As it was previously found for the first three species that a signifi-cant positive correlation occurred between Hg concentrations andsize or weight, the appropriate factor was used as covariable in theANCOVA performed to test the differences among sites.

For all the four species, Hg concentrations in the edible tissuesdisplayed significant differences among sampling sites (pANCOVA or

ANOVA always <0.001). Post hoc Tukey test also showed that for allspecies, the highest Hg concentrations were always measured inGrande Rade. With respect to A. scapha and G. tumidum, relativelylow Hg values were found for individuals collected in Ouano Baysites (ranging from 73 to 492 ng g�1 dry wt). In contrast, the fishcollected in the latter bay displayed quite high Hg concentration(from 591 to 1338 ng g�1 dry wt in muscles). Koutio, Maa and SainteMarie bays were generally characterised by intermediate values forthe three bivalves, and by the lowest values for the fish (Table 4).Regarding the clam G. tumidum, Hg concentrations were the lowestin Unia, whereas they were relatively high in Tomo (154� 38 and497�124 ng g�1 dry wt, respectively). Nevertheless, values inTomo were significantly lower than those measured in Grande Rade

Table 2Latitude and longitude coordinates of the sampling stations in the New CaledoniaLagoon

Sampling site Geographical coordinates

Latitude Longitude

Northern Lagoon 19�57 S 163�58 EBassin La Foa 21�45 S 165�43 EOuano Bay S1 21�51 S 165�50 EOuano Bay S2 21�49 S 165�46 EOuano Bay S3 21�52 S 165�49 ETomo 21�57 S 166�04 EMaa Bay 22�12 S 166�19 EDumbea Bay 22�12 S 166�24 EKoutio Bay 22�13 S 166�25 EGrande Rade 22�16 S 166�25 EGrande Rade-SLN 22�15 S 166�26 ENoumea harbour 22�17 S 166�25 ESainte Marie Bay 22�18 S 166�28 EUnia 21�55 S 166�38 ESouthern Lagoon 22�25 S 166�53 E

(Table 4). At a smaller geographic scale, possible difference betweenGrande Rade and Grande Rade-SLN and among the sub-stations inOuano Bay (S1, S2, and S3) could be investigated for two out of thefour species: A. scapha and G. tumidum. For G. tumidum, Hg con-centrations were significantly higher (p< 0.001) in Grande Radethan in Grande Rade-SLN, whereas no significant difference wasdetected for A. scapha. In Ouano Bay, no significant difference couldbe detected among the three sampling sites (Table 4).

3.4. Risk assessment for Human consumers

The ‘‘Maximum Safe Weekly Consumption’’ (MSWC, in kg wetwt) of edible flesh was estimated taking into account the ‘‘Pro-visional Tolerable Weekly Intake’’ (PTWI) recommended by JECFA(WHO, 2003) for total Hg and methyl-Hg. MSWC values are given ona species basis in Table 3 (for the species collected in more than onesite, the MSWC is thus an average value taking into account thedifferent sampling sites), and on a site-specific basis in Table 4 (forthe four species that were collected in sufficient quantity in severalsites). The amount of shrimp muscle (L. stylirostris) from Bassin deLa Foa (i.e. a shrimp farm) which should be eaten by a 60-kg personto reach the PTWI for total Hg (5000 ng kg�1 wk�1) would be ca.7.5 kg. For the oyster C. gigas from Dumbea Bay, this amount wouldbe about 13.5 kg of oyster flesh. Regarding cephalopods from thesouthern and northern lagoon, the MSWC for total Hg ranged from10 kg of edible flesh (muscle) for the squid S. lessoniana to36.7 kg wk�1 for the cuttlefish S. latimanus (Table 3). With respect tofish consumption, the MSWC for total Hg was between ca. 1 kg offlesh for some carnivorous species from Grande Rade (Plecto-rhinchus chaetodonoides, Plectorhinchus flavomaculatus) to morethan 32 kg wk�1 for some herbivorous species coming from OuanoBay (Naso unicornis, Scarus microrhinos, Scarus rivulatus, Scarusschlegeli) (Table 3). In Grande Rade, a 60-kg person should eat6.4 kg, 2.1 kg and 920 g wet wt wk�1 of flesh of C. radula, A. scaphaand G. tumidum, respectively, to reach the PTWI in total Hg (Table 4).In the same site, the MSWC for total Hg was 830 g of edible flesh ofthe P. leopardus fish. In all cases, MSWCs for methyl-Hg were farlower than those computed for total Hg, especially in fish (Tables 3

Table 3Total Hg concentrations (T-Hg; mean� SD, ng g�1 dry wt) in the edible tissues of the species collected, and estimate of the ‘‘Maximum Safe Weekly Consumption’’ (MSWC) fortotal Hg (T-Hg) and methyl-Hg (Me-Hg) (kg of flesh wet wt)

Species Tissue n T-Hg concentration (ng g�1 dry wt) MSCW (kg)

Mean� SD (Range) T-Hg Me-Hg

CrustaceansLitopenaeus stylirostris Muscle 5 201� 62 128–297 7.5 3.2

CephalopodsSepioteuthis lessoniana Muscle 9 150� 36 113–218 10 4Octopus cyanea Muscle 7 90� 23 69–139 16.7 6.7Sepia latimanus Muscle 1 41 – 36.7 14.7

BivalvesComptopallium radula Muscleþ gonad 24 62� 20 30–108 24.3 –C. radula Whole soft partsa 24 150� 53 84–261 10 –Mimachlamys gloriosa Muscleþ gonad 10 70� 12 55–91 21.5 –M. gloriosa Whole soft partsa 10 107� 18 90–142 14 –Anadara scapha Whole soft parts 28 297� 226 73–1015 5.1 –Crassostrea gigas Whole soft parts 5 111� 15 92–105 13.5 –Gafrarium tumidum Whole soft parts 89 418� 466 104–2531 3.6 –Periglypta chemnitzi Whole soft parts 10 407� 222 244–890 3.7 –

FishNaso unicornis Muscle 1 32 – 38 12.1Plectorhinchus albovittatus Muscle 1 650 – 1.8 0.59P. chaetodonoides Muscle 1 1200 – 1 0.32P. flavomaculatus Muscle 1 1173 – 1 0.33Diagramma pictum Muscle 1 329 – 3.6 1.2Kyphosus vaigiensis Muscle 5 55� 22 36–92 22 7Bodianus perditio Muscle 1 1020 – 1.2 0.38Cheilinus chlorourus Muscle 1 362 – 3.3 1.1Lethrinus laticaudis Muscle 3 165� 55 125–228 7.3 2.3Monotaxis grandoculis Muscle 2 99� 18 87–112 12.1 3.9Lutjanus argentimaculatus Muscle 5 430� 325 403–994 2.8 0.89L. monostigma Muscle 2 51� 23 35–67 23.6 7.5Cymbacephalus beauforti Muscle 2 1022� 62 979–1066 1.2 0.38Priacanthus hamrur Muscle 7 188� 58 113–258 6.4 2Scarus ghobban Muscle 1 53 – 22.7 7.3S. microrhinos Muscle 1 30 – 39.5 12.6S. rivulatus Muscle 1 26 – 46.1 14.8S. schlegeli Muscle 1 37 – 32.2 10.3Plectropomus leopardus Muscle 21 738� 576 173–2063 1.6 0.52Epinephelus coeruleopunctatus Muscle 1 439 – 2.7 0.87E. maculatus Muscle 1 389 – 3.1 0.99Acanthopagrus berda Muscle 2 368� 168 249–486 3.2 1

a Reconstructed data.

T. Chouvelon et al. / Environmental Pollution 157 (2009) 331–340336

and 4). For example, a 60-kg person should eat about 260 g wet wtof flesh of P. leopardus from Grande Rade to reach the PTWI inmethyl-Hg (vs. the aforementioned 830 g for total Hg).

4. Discussion

The whole dataset of this study provides substantial baselineinformation of major importance for future investigations on Hgcontamination status in New Caledonia as well as for the evaluationof the risk related to local seafood consumption. Indeed, despite thewell-known toxicity of Hg to marine organisms and its potentialeffects on Human health, this metal has been poorly investigated inthis tropical zone, with only one published study reporting Hgmeasurements in tissues of two marine mammals (Bustamanteet al., 2003). So far, there are no published data on Hg levels ininvertebrate or vertebrate species used as food by local populations,whether they are fished from the wild or cultured in artificial(shrimp) or natural (oysters) environments.

In species for which several tissues were examined, the lowestconcentrations were almost always recorded in the muscles (seeFigs 2–4). In cephalopods, the digestive gland and the branchialhearts are actually known to be major sites for concentratingmetals, among which Hg, compared to muscles (e.g. Miramand andBentley, 1992; Bustamante et al., 2000, 2006). The digestive gland isgenerally supposed to play a major role in the storage and

detoxification of many essential and non-essential elements suchas Ag, Cd, Cu or Zn (Miramand and Bentley, 1992; Bustamante et al.,2000, 2002, 2006). Concerning the Pectinidae family, the digestivegland and the kidneys have been identified for their storage anddetoxification function for many elements (e.g. Bustamante andMiramand, 2004, 2005; Metian et al., 2008), especially in the caseof Hg (Metian et al., submitted for publication). This explains thehigher Hg values that we measured in the ‘‘remaining soft tissues’’body compartment of C. radula and M. gloriosa (Fig. 4) as thiscompartment comprises all tissues and organs other than the ad-ductor muscle and the gonad (i.e. it includes both the kidneys andthe digestive gland). Finally, in fish, hepatic levels of Hg are gen-erally reported to be higher than the muscular ones (e.g. Storelliet al., 2005; Kojadinovic et al., 2007), probably due to the deme-thylation process that has been suggested to occur in fish liver, asindicated by the occurrence of almost only inorganic Hg in thisorgan (Oliveira Ribeiro et al., 1996; Maury-Brachet et al., 2006). Inour study, liver was similarly found to be the major organ for Hgstorage in fish. Pertaining to edible tissues, Hg concentrationsvaried greatly in some bivalves collected from different sites in theNew Caledonia lagoon. Hg levels ranged from 73 to 1015 and from104 to 2531 ng g�1 dry wt in the whole soft parts of A. scapha and G.tumidum, respectively. In fish, P. leopardus, which was also collectedin several sites, displayed muscular Hg concentrations ranging from173 to 2063 ng g�1 dry wt. On the other hand, in some species of

Table 4Comparison of total Hg concentrations (T-Hg; mean� SD, ng g�1 dry wt) in the whole soft parts of the bivalves Comptopallium radula, Anadara scapha and Gafrarium tumidumand in the muscle of the fish Plectropomus leopardus collected from different sites in New Caledonia

Species Site n T-Hg concentration(ng g�1 dry wt)

Group MSCW (kg)

Mean� SD Range 1 2 3 4 T-Hg Me-Hg

Bivalves (whole soft parts)Anadara scapha One-way ANOVA and Tukey testa

Ouano Bay S2 5 103� 39 73–169 j 14.6 –Ouano Bay S1 3 146� 20 130–169 j j 10.3 –Koutio Bay 5 230� 111 129–354 j j j 6.5 –Maa Bay 5 304� 165 164–588 j j 4.9 –Sainte Marie Bay 5 337� 145 224–279 j j 4.4 –Grande Rade-SLN 3 515� 290 184–724 j j 2.9 –Grande Rade 2 723� 414 430–1015 j 2.1 –

Comptopallium radula ANCOVA (cov.: length) and Tukey testa

Noumea harbour 3 99� 13 84–108 j 15.2 –Maa Bay 6 116� 19 93–144 j 12.9 –Sainte Marie Bay 7 125� 16 103–145 j 12.0 –Koutio Bay 4 199� 29 173–231 j 7.6 –Grande Rade 4 234� 24 203–261 j 6.4 –

Gafrarium tumidum ANCOVA (cov.: weight) and Tukey testa

Unia 10 154� 38 111–224 j 9.7 –Ouano Bay S3 10 193� 45 146–281 j j 7.8 –Ouano Bay S2 10 197� 68 106–295 j j 7.6 –Grande Rade-SLN 10 253� 107 143–495 j j 5.9 –Koutio Bay 10 299� 143 104–614 j 5.0 –Ouano Bay S1 10 310� 98 198–492 j j 4.8 –Maa Bay 10 344� 194 113–741 j j 4.4 –Tomo 10 497� 124 291–658 j 3.0 –Grande Rade 9 1633� 565 894–2531 j 0.92 –

Fish (muscle)Plectropomus leopardus ANCOVA (cov.: length) and Tukey testa

Koutio Bay 4 336� 49 286–380 j 3.6 1.1Maa Bay 4 353� 82 245–445 j 3.4 1.1Sainte Marie Bay 3 629� 280 436–950 j j 1.9 0.61Ouano Bay S3 4 856� 330 591–1338 j j 1.4 0.45Grande Rade 5 1451� 697 449–2063 j 0.83 0.26

MSWC is the estimate of the ‘‘Maximum Safe Weekly Consumption’’ for total Hg (T-Hg) and methyl-Hg (Me-Hg) (kg of flesh wet wt) from each site.a According to the results of ANOVA or ANCOVA followed by Tukey test, means which do not differ significantly within each species are joined by vertical bars (j).

T. Chouvelon et al. / Environmental Pollution 157 (2009) 331–340 337

fish (i.e. Scarus sp., Lutjanus monostigma, K. vaigiensis), Hg concen-trations were very low (<60 ng g�1 dry wt) for most of the in-dividuals collected (Table 3). Similarly, Hg concentrations measuredin the cultured species did not exceed 105 and 297 ng g�1 dry wt forthe oyster C. gigas and the shrimp L. stylirostris, respectively.

Comparison with other tropical areas (i.e. NE coast of Australia,Thailand, and Gulf of Oman) indicates that the total Hg concen-trations reported in similar organisms were generally below themean values that we obtained in this study. Indeed, total Hg inbivalves from Thailand – the clam Paphia undulata, the blood cockle

Fig. 2. Relationship between total Hg concentrations (T-Hg; ng g�1 dry wt) in liver andmuscle in fish species from New Caledonia.

Anadara granosa and the oyster Crassostrea commercialis – neverexceeded 100 ng g�1 dry wt (Phillips and Muttarasin, 1985)whereas in the Gulf of Oman, Hg concentrations were below315 ng g�1 dry wt for the clam Circentia callipyga, the spiny oysterSpondylus sp. and the oyster Saccostrea cucullata (de Mora et al.,2004). Regarding fish, relatively low concentrations were reportedfor species from the NE coast of Australia with Hg levels rangingfrom 120 to 760 ng g�1 dry wt in muscle of Acanthopagrus berda,Cymbacephalus nematophthalmus, Lutjanus carponotatus, Lutjanussanguineus and Plectorhinchus flavimaculatus (Denton and Breck,

0

100

200

300

400

Octopus cyanea Sepioteuthis lessoniana Sepia latimanus

T-H

g co

ncen

tratio

n (n

g g

-1 d

wt)

n = 7 n = 9 n = 1

Digestive glandBranchial heartsMuscle

Fig. 3. Total Hg concentrations (T-Hg; mean� SD, ng g�1 dry wt) in the digestive gland,branchial hearts and muscle of three cephalopod species from New Caledonia.

0

100

200

300

400

Comptopallium radula Mimachlamys gloriosa

T-H

g co

ncen

tratio

n (n

g g

-1 d

wt)

n = 24 n = 10

Muscle + GonadRemaining tissuesWhole soft parts*

Fig. 4. Total Hg concentrations (T-Hg; mean� SD, ng g�1 dry wt) in themuscleþ gonad and in the remaining tissues of two scallop species from NewCaledonia. *Reconstructed data.

T. Chouvelon et al. / Environmental Pollution 157 (2009) 331–340338

1981). However, de Mora et al. (2004) found relatively high valuesin fish from the Gulf of Oman, i.e. ranging from 343 to 522 and from498 to 2350 ng g�1 dry wt in Lethrinus nebulosus and Epinepheluscoicoides, respectively. Also, Denton and Breck (1981) reportedvalues from 250 to 1250 ng Hg g�1 dry wt in muscles of cephalo-pods (Loligo sp. and Sepia sp.), and an Hg concentration of ca.1000 ng g�1 dry wt for the tumid venus G. tumidum. Moreover, inchronically polluted, non-tropical areas (i.e. Western Limfjord,Denmark and Lavaca Bay, Texas, USA), Hg concentrations rangingfrom 414 to 2517 ng g�1 dry wt were reported in the cockle Cardiumsp. and from 430 to 10 100 ng g�1 dry wt in several other species ofbivalves (Mohlenberg and Riisgard, 1988; Locarnini and Presley,1996). In our study, some values recorded in the edible A. scaphaand G. tumidum from Grande Rade were thus on the same order ofmagnitude than those reported in well-known Hg-contaminatedareas. Hence, along with our data on Hg levels in fish muscles, theformer comparisons indicate that New Caledonia may suffer Hgcontamination to some extent and that additional studies would benecessary for further assessing the contamination status in Hg ofthe New Caledonia lagoon. Nonetheless, the assessment of the riskfor Human consuming marine products from this lagoon appearedof particular concern, particularly for the Grande Rade area.

T-H

g co

ncen

tratio

n (n

g g

-1 d

wt)

GS PI PISM PSM

0

500

1000

1500

2000 Non-outlier Max.Non-outlier Min.3rd quartile

Mean1st quartileOutlier

Fig. 5. Comparison of total Hg concentrations (T-Hg, ng g�1 dry wt) in the muscles offish from New Caledonia according to their trophic level. GS¼ grazer/scavenger(n¼ 10); PI¼ predator of invertebrates (n¼ 5); PISM¼ predator of invertebrates andsmall fish (n¼ 23); and PSM¼ predator of small fish (n¼ 24).

From a global health standpoint, Hg concentrations measured inthe investigated organisms of this study did not reveal excessiverisk for Humans consuming seafood in New Caledonia. Indeed,even in the most contaminated site (Grande Rade), the amount ofbivalve flesh to be eaten on a weekly basis by a 60-kg person fora given species to reach the PTWI for total Hg was relatively ele-vated: 920 g, 2.1 kg and 6.4 kg wet wt for G. tumidum, A. scapha andC. radula, respectively (Table 4). Also, regarding the fish P. leopardusin the same location, a 60-kg person should eat more than 830 g offish muscle to exceed the PTWI in total Hg. Hence, except in localHuman groups where sea products constitute the main food source,the former quantities of fish or bivalves are not very likely to bereached over a 1-week period, as far as total Hg is considered.However, if we assume that ca. 100% of total Hg in fish muscle isoccurring as methyl-Hg as it is typically the case in this group (e.g.Bloom, 1992; Holsbeek et al., 1997), the PTWI in methyl-Hg wouldbe reached with the consumption of approximately 260 g wet wt ofedible flesh of P. leopardus from Grande Rade. Such a low quantity offlesh can be easily consumed during a single meal. In New Cale-donia, fish consumption is particularly high, varying between 23and 50 kg/inhabitant/year, i.e. between 0.5 and 1 kg wk�1 (Labrosseet al., 2006). Moreover, the ‘‘Maximum Safe Weekly Consumption’’(MSWC) for methyl-Hg assessed for the other fish species collectedin Grande Rade (i.e. P. chaetodonoides, P. flavomaculatus and Bod-ianus perditio) were similarly quite low and always below 400 g wetwt of flesh per week for a 60-kg consumer (Table 3). Although theselow MSWC values are of concern, they were only found in the singlearea of Grande Rade.

As for other metals, Hg concentrations in marine organisms mayvary with biological and environmental factors such as age (size/weight), trophic level, lifestyle and geographical origin (e.g. Mon-teiro et al., 1992; Otchere et al., 2003; Burger et al., 2007; Zizeket al., 2007). Size has been reported as a factor to which Hg con-centrations are well correlated in fish (Monteiro and Lopes, 1990;Mathieson and McLusky, 1995; Adams, 2004; Kojadinovic et al.,2006), due to the continuous bioaccumulation of Hg in musculartissues all along their life (Braune, 1987; Burger and Gochfeld,2007). However, among the fish species examined in this study,only P. leopardus displayed a positive correlation with length andweight. The lack of positive correlation for the other fish in-vestigated may be due to the small number of individuals withinthose species, or, alternatively to the fact that the sampled in-dividuals were of about the same age (Braune, 1987; Mathieson andMcLusky, 1995).

In cephalopods, it is also generally admitted that Hg concen-trations are positively correlated with size (Rossi et al., 1993; Bus-tamante et al., 2006; Pierce et al., 2008). However, no correlationwas found either for S. lessoniana or for O. cyanea, probably becauseof the too narrow size range of the individuals examined.

Hg concentrations were found to differ among species, andnotably among species belonging to the same taxonomic group.Interestingly, for a given sampling location (Grande Rade, KoutioBay, Maa Bay), the three species of bivalves (C. radula, A. scapha andG. tumidum) did not bioaccumulate Hg to the same extent. Theincreasing order of concentrations (C. radula< A. scapha<G.tumidum) always occurred in each site, the clam and the cockleconcentrating Hg much more than the scallop. These three specieswere of very different sizes: average length was 77�14 mm for C.radula, 58� 9 mm for A. scapha and 38� 5 mm for G. tumidum. It iswell established that metal uptake efficiency can decrease withorganism size increase (e.g. Swaileh and Adelung, 1995; Warnauet al., 1995; Hedouin et al., 2006). This trend is generally explainedby the decreasing surface/volume ratio of an organism with in-creasing body size, the decreasing metabolic activity in older in-dividuals, and/or by the dilution of the metal in larger organisms(e.g. Braune, 1987; Swaileh and Adelung, 1995; Warnau et al., 1995).

T. Chouvelon et al. / Environmental Pollution 157 (2009) 331–340 339

Moreover, different species of bivalves may have different diet andfiltration rates. Although further studies are needed to better ex-plain the differences in Hg bioaccumulation among marine bivalvesfrom New Caledonia, it is effectively well known that scallops havea quite elevated filtration rate compared to other bivalves(Meyhofer, 1985). As they displayed lower values of Hg than otherbivalves in this study, it could be suggested that diet compositioncould be a major factor in Hg bioaccumulation in bivalves. This is inaccordance with the few evidences available in the related litera-ture (e.g. Wang and Wong, 2003; Metian et al., submitted forpublication). This apparently also holds true in fish, for which thetrophic level clearly appeared to affect Hg concentrations in theinvestigated species (see Fig. 5). Indeed, grazers/scavengers dis-played the lowest values of Hg in their muscular tissues, whereaspredators of small fish displayed the highest ones. These observa-tions are in good agreement with the Hg biomagnification processoccurring in marine food webs (e.g. Burger et al., 2007; Zizek et al.,2007).

Species, size or weight and trophic level as factors of variationfor Hg concentrations were also well illustrated by the case of thethree fish species collected in Ouano Bay S3. Indeed, they displayedsignificant differences in their mean muscle values of total Hg,according to the following order: K. vaigiensis< P. hamrur< P.leopardus. Actually, K. vaigiensis and P. hamrur are similar in size, butthe former is a grazer whereas the second is a predator of in-vertebrates and small fish; as for P. leopardus, it is much heavier andis a predator of small fish, and displayed thus, as could be expected,the highest Hg concentrations among the three species.

Finally, there were clear differences in Hg concentrations amongsites for a given species, the site factor explaining between 68 and85% of the variability in the performed ANCOVA. In the three bi-valve species A. scapha, C. radula and G. tumidum, and in the fish P.leopardus, individuals from Grande Rade always displayed thehighest concentrations in Hg. Considering the concentrations ofseveral metals in the sediment and in the clam G. tumidum, GrandeRade is regarded as a highly contaminated station (Hedouin et al., inpress). As this area is subjected to industrial inputs, it was notsurprising to find high concentrations of Hg in the organisms livingthere. However, there was a spatial heterogeneity for bivalves fromthe Grande Rade area, with Hg concentrations in G. tumidum fromGrande Rade-SLN that were surprisingly much lower than thatfrom Grande Rade (individuals from both locations were of similarsize). Very intense exposure to Ni and to Ni-ore co-occurring metalsin Grande Rade-SLN lead to very high concentrations on these el-ements in the tissues of G. tumidum (>10 mg Cr g�1, >90 mg Ni g�1,>300 mg Mn g�1; Hedouin et al., in press) which could interferewith other elements such as Hg. However, the lack of similar trendsin Hg concentrations for the other bivalve also collected in bothsites (A. scapha) does not support this hypothesis. Further in-vestigations are thus necessary to explain this peculiar observationin G. tumidum.

Acknowledgments

The authors are grateful to O. Pringault (IRD-Noumea) for hishelp in organising the field work, to Captain M. Clarque (IRD-Noumea), A.M. Rodriguez y Baena (CIESM, Principality of Monaco)and F. Pedraza-Diaz (University of La Rochelle, France) for skilfulassistance in fish sampling and to and the IRD Diver Team forproviding access to their diving equipment. MW is an HonorarySenior Research Associate of the National Fund for Scientific Re-search (NFSR, Belgium) and holds a 2008 Invited Expert position atLIENSs (UMR 6250, CNRS-Universite de La Rochelle), supported bythe Conseil Regional de Poitou-Charentes. The IAEA is gratefulfor the support provided to its Marine Environment Laboratoriesby the Government of the Principality of Monaco. This work was

supported by the IAEA, the PNEC programme (Chantier Nouvelle-Caledonie) and the IRD.

References

Adams, D.H., 2004. Total mercury levels in tunas from offshore waters of the FloridaAtlantic coast. Marine Pollution Bulletin 49, 659–667.

Ambastian, P., Fernex, F., Bernat, M., Parron, C., Lecolle, J., 1997. High metal inputs toclosed seas: the New Caledonia lagoon. Journal of Geochemistry Exploration 59,59–74.

Bloom, N.S., 1992. On the chemical form of mercury in edible fish and marine in-vertebrate tissue. Canadian Journal of Fisheries and Aquatic Sciences 49,1010–1017.

Braune, B.M., 1987. Mercury accumulation in relation to size and age of AtlanticHerring (Clupea harengus harengus) from the Southwestern Bay of Fundy,Canada. Archives Environmental Contamination and Toxicology 16, 311–320.

Burger, J., Gochfeld, M., 2007. Risk to consumers from mercury in Pacific cod (Gadusmacrocephalus) from the Aleutians: fish age and size effects. EnvironmentalResearch 105, 276–284.

Burger, J., Gochfeld, M., Jeitner, C., Burke, S., Stamm, T., Snigaroff, R., Snigaroff, D.,Patrick, R., Wetson, J., 2007. Mercury levels and potential risk from sub-sistence foods from the Aleutians. Science of the Total Environment 384,93–105.

Bustamante, P., Cosson, R.P., Gallien, I., Caurant, F., Miramand, P., 2002. Cadmiumdetoxification processes in the digestive gland of cephalopods in relation toaccumulated cadmium concentrations. Marine Environmental Research 53,227–241.

Bustamante, P., Garrigue, C., Breau, L., Caurant, F., Dabin, W., Greaves, J.,Dodemont, R., 2003. Trace elements in two odontocete species (Kogia brevicepsand Globicephala macrorhynchus) stranded in New Caledonia (South Pacific).Environmental Pollution 124, 263–271.

Bustamante, P., Grigioni, S., Boucher-Rodoni, R., Caurant, F., Miramand, P., 2000.Bioaccumulation of 12 trace elements in the tissues of the nautilus Nautilusmacromphalus from New Caledonia. Marine Pollution Bulletin 40, 688–696.

Bustamante, P., Lahaye, V., Durnez, C., Churlaud, C., Caurant, F., 2006. Total andorganic Hg concentrations in cephalopods from the North Eastern Atlanticwaters: influence of geographical origin and feeding ecology. Science of theTotal Environment 368, 585–596.

Bustamante, P., Miramand, P., 2004. Interspecific and geographical variations oftrace element concentrations in Pectinidae from European waters. Chemo-sphere 57, 1355–1362.

Bustamante, P., Miramand, P., 2005. Subcellular and body distributions of 17 traceelements in the variegated scallop Chlamys varia from the French coast of theBay of Biscay. Science of the Total Environment 337, 59–73.

Buzina, R., Suboticanec, K., Vukusic, J., Sapunar, J., Antonic, K., Zorica, M., 1989. Effectof industrial pollution on seafood content and dietary intake of total andmethylmercury. Science of the Total Environment 78, 45–57.

Cossa, D., Thibaud, Y., Romeo, M., Gnassia-Barelli, M., 1990. Le mercure en milieumarin. Biogeochimie et ecotoxicologie. Rapports Scientifiques et Techniques del’IFREMER 19 Brest.

Denton, G.R.W., Breck, W.G., 1981. Mercury in tropical marine organisms from NorthQueensland. Marine Pollution Bulletin 12, 116–121.

Eisler, R., 1987. Mercury Hazards to Fish, Wildlife, and Invertebrates: a SynopticReview. U.S. Fish and Wildlife Service Biological Report 85 (1.10).

Hedouin, L., Metian, M., Teyssie, J.-L., Fowler, S.W., Fichez, R., Warnau, M., 2006.Allometric relationships in the bioconcentration of heavy metals by the edibletropical clam Gafrarium tumidum. Science of the Total Environment 366,154–163.

Hedouin, L., Pringault, O., Metian, M., Bustamante, P., Warnau, M., 2007. Nickel bi-oaccumulation in bivalves from the New Caledonia lagoon: seawater and foodexposure. Chemosphere 66, 1449–1457.

Hedouin, L., Bustamante, P., Churlaud, C., Pringault, O., Fichez, R., Warnau, M. Trendsin concentrations of selected metalloid and metals in two bivalves from the SWlagoon of New Caledonia. Ecotoxicology and Environmental Safety, doi:10.1016/j.ecoenv.2008.04.004, in press.

Holsbeek, L., Das, K., Joiris, C.R., 1997. Mercury speciation and accumulation inBangladesh freshwater and anadromous fish. Science of the Total Environment198, 201–210.

Hylander, L.D., 2001. Global mercury pollution and its expected decrease aftera mercury trade ban. Water, Air, and Soil Pollution 125, 331–344.

Kawaguchi, T., Porter, D., Bushek, D., Jones, B., 1999. Mercury in the American oyster(Crassostrea virginica) in South Carolina, USA, and public health concerns. Ma-rine Pollution Bulletin 38, 324–327.

Kojadinovic, J., Potier, M., Le Corre, M., Cosson, R.P., Bustamante, P., 2006. Mercurycontent in commercial pelagic fish and its risk assessment in the Western In-dian Ocean. Science of the Total Environment 366, 688–700.

Kojadinovic, J., Potier, M., Le Corre, M., Cosson, R.P., Bustamante, P., 2007. Bio-accumulation of trace elements in pelagic fish from the Western Indian Ocean.Environmental Pollution 146, 548–566.

Labrosse, P., Ferraris, J., Letourneur, Y., 2006. Assessing the sustainability of sub-sistence fisheries in the Pacific: the use of data on fish consumption. OceanCoastline Management 49, 203–221.

Labrosse, P., Fichez, R., Farman, R., Adams, T., 2000. New Caledonia. In: Sheppard, C.R.C. (Ed.), Seas at the Millennium: an Environmental Evaluation. Elsevier Sci-ence, Amsterdam, pp. 723–736.

T. Chouvelon et al. / Environmental Pollution 157 (2009) 331–340340

Locarnini, S.J.P., Presley, B.J., 1996. Mercury concentrations in benthic organismsfrom a contaminated estuary. Marine Environmental Research 41, 225–239.

Maury-Brachet, R., Durrieu, G., Dominique, Y., Boudou, A., 2006. Mercury distri-bution in fish organs and food regimes: significant relationships from twelvespecies collected in French Guiana (Amazonian basin). Science of the TotalEnvironment 368, 262–270.

Mathieson, S., McLusky, D.S., 1995. Inter-species variation of mercury in skeletalmuscle of five fish species from inshore waters of the Firth of Clyde, Scotland.Marine Pollution Bulletin 30, 283–286.

Metian, M., Bustamante, P., Hedouin, L., Warnau, M., 2008. Accumulation of traceelements in the tropical scallop Comptopallium radula from coral reefs in NewCaledonia. Environmental Pollution 152, 543–552.

Metian, M., Warnau, M., Cosson, R.P., Bustamante, P. Bioaccumulation of Hg anddetoxification processes in the king scallop Pecten maximus: field and labora-tory investigations, submitted for publication.

Meyhofer, E., 1985. Comparative pumping rates in suspension-feeding bivalves.Marine Biology 85, 137–142.

Miramand, P., Bentley, D., 1992. Concentration and distribution of heavy metals intissues of two cephalopods, Eledone cirrhosa and Sepia officinalis, from theFrench coast of the English Channel. Marine Biology 114, 407–414.

de Mora, S., Fowler, S.W., Wyse, E., Azemard, S., 2004. Distribution of heavy metalsin marine bivalves, fish and coastal sediments in the Gula of Oman. MarinePollution Bulletin 49, 410–424.

Mohlenberg, F., Riisgard, H.U., 1988. Partitioning of inorganic and organic Mercuryin cockles Cardium edule (L.) and C. glaucum (Bruguiere) from a chronicallypolluted area: influence of size and age. Environmental Pollution 55, 137–148.

Monniot, F., Martoja, R., Monniot, C., 1994. Cellular sites of iron and nickel accu-mulation in ascidians related to the naturally and anthropic enriched NewCaledonian environment. Annales de l’Institut Oceanographique 70, 205–216.

Monteiro, L.R., Lopes, H.D., 1990. Mercury content of Swordfish, Xiphias gladius, inrelation to length, weight, age and sex. Marine Pollution Bulletin 21, 293–296.

Monteiro, L.R., Porteiro, F.M., Gonçalves, J.M., 1992. Inter- and intra-specific varia-tion of mercury levels in muscle of cephalopods from the Azores. Archipelago10, 13–22.

Oliveira Ribeiro, C.A., Guimaraes, J.R.D., Pfeiffer, W.C., 1996. Accumulation and dis-tribution of inorganic mercury in a tropical fish (Trichomycterus zonatus). Eco-toxicology and Environmental Safety 34, 190–195.

Otchere, F.A., Joiris, C., Holsbeek, L., 2003. Mercury in the bivalves Anadara (Senilis)senilis, Perna perna and Crassostrea tulipa from Ghana. Science of the TotalEnvironment 304, 369–375.

Phillips, D.J.H., Muttarasin, K., 1985. Trace metals in bivalve molluscs from Thailand.Marine Environmental Research 15, 215–234.

Pierce, G.J., Stowasser, G., Hastie, L.C., Bustamante, P., 2008. Geographic, seasonaland ontogenetic variation in cadmium and mercury concentrations in squid

(Cephalopoda: Teuthoidea) from UK waters. Ecotoxicology and EnvironmentalSafety 70, 422–432.

Riisgard, H.U., Famme, P., 1986. Accumulation of inorganic and organic mercury inshrimp, Crangon crangon. Marine Pollution Bulletin 17, 255–257.

Rossi, A., Pellegrini, D., Belcari, P., Barghigiani, C., 1993. Mercury in Eledone cirrhosafrom the Northern Tyrrhenian Sea: contents and relations with life cycle. Ma-rine Pollution Bulletin 26, 683–686.

Salomons, W., de Rooij, N.M., Derdijk, H., Bril, J., 1987. Sediments as a source forcontaminants? Hydrobiologia 149, 13–30.

Storelli, M.M., Giacominelli, R., Storelli, A., Marcotrigiano, G.O., 2005. Accumulationof mercury, cadmium, lead and arsenic in Swordfish and Bluefin tuna from theMediterranean Sea: a comparative study. Marine Pollution Bulletin 44,281–288.

Svensson, B.G., Schutz, A., Nilsson, A., Akesson, I., Akesson, B., Skerfving, S., 1992.Fish as a source of exposure to mercury and selenium. Science of the TotalEnvironment 126, 61–74.

Swaileh, K.M., Adelung, D., 1995. Effect of body size and season on the concentra-tions of Cu, Cd, Pb and Zn in Diastylis rathkei (Kroyer) (Crustacea: Cumacea)from Kiel Bay, Western Baltic. Marine Pollution Bulletin 31, 103–107.

Trombini, C., Fabbri, D., Lombardo, M., Basura, I., Zavoli, E., Horvat, M., 2003. Mer-cury and methylmercury contamination in surficial sediments and clams ofa coastal lagoon (Pialassa Baiona, Ravenna, Italy). Continental Shelf Research 23,1821–1831.

Wang, W.-X., Wong, R.C.K., 2003. Combined effects of food quantity and quality onCd, Cr, and Zn assimilation to the green mussel, Perna viridis. Journal of Ex-perimental Marine Biology and Ecology 290, 49–69.

Warnau, M., Biondo, R., Temara, A., Bouquegneau, J.M., Jangoux, M., Dubois, P., 1998.Distribution of heavy metals in the echinoid Paracentrotus lividus (Lmk) fromthe Mediterranean Posidonia oceanica ecosystem: seasonal and geographicalvariations. Journal of Sea Research 39, 267–280.

Warnau, M., Ledent, G., Temara, A., Alva, V., Jangoux, M., Dubois, P., 1995. Allometryof heavy metal bioconcentration in the echinoid Paracentrotus lividus (Echino-dermata). Archives of Environmental Contamination and Toxicology 29,393–399.

Warnau, M., Teyssie, J.-L., Fowler, S.W., 1996. Biokinetics of selected heavy metalsand radionuclides in two marine macrophytes: the seagrass Posidonia oceanicaand the alga Caulerpa taxifolia. Marine Environmental Research 41, 343–362.

WHO, 1990. Methylmercury. Environmental Health Criteria 101, 1–80.WHO, 2003. Joint FAO/WHO Expert Committee on Food Additives and Contami-

nants, Sixty-first Meeting. Summary and Conclusions Available from: ftp://ftp.fao.org/es/esn/jecfa/jecfa61sc.pdf.

Zizek, S., Horvat, M., Gibicar, D., Fajon, V., Toman, M.J., 2007. Bioaccumulation ofmercury in benthic communities of a river ecosystem affected by mercurymining. Science of the Total Environment 377, 407–415.